---
license: apache-2.0
---

<div align="center">
  <h1>X2Edit</h1>
<a href='https://arxiv.org/abs/2508.07607'><img src='https://img.shields.io/badge/arXiv-2508.07607-b31b1b.svg'></a> &nbsp;
<a href='https://huggingface.co/datasets/OPPOer/X2Edit-Dataset'><img src='https://img.shields.io/badge/🤗%20HuggingFace-X2Edit Dataset-ffd21f.svg'></a>
<a href='https://huggingface.co/OPPOer/X2Edit'><img src='https://img.shields.io/badge/🤗%20HuggingFace-X2Edit-ffd21f.svg'></a>
<a href='https://www.modelscope.cn/datasets/AIGCer-OPPO/X2Edit-Dataset'><img src='https://img.shields.io/badge/🤖%20ModelScope-X2Edit Dataset-purple.svg'></a>
</div>

## Environment

Prepare the environment, install the required libraries:

```shell
$ git clone https://github.com/OPPO-Mente-Lab/X2Edit.git
$ cd X2Edit
$ conda create --name X2Edit python==3.11
$ conda activate X2Edit
$ pip install -r requirements.txt
```

## Inference
We provides inference scripts for editing images with resolutions of **1024** and **512**. In addition, we can choose the base model of X2Edit, including **[FLUX.1-Krea](https://huggingface.co/black-forest-labs/FLUX.1-Krea-dev)**, **[FLUX.1-dev](https://huggingface.co/black-forest-labs/FLUX.1-dev)**, **[FLUX.1-schnell](https://huggingface.co/black-forest-labs/FLUX.1-schnell)**, **[PixelWave](https://huggingface.co/mikeyandfriends/PixelWave_FLUX.1-dev_03)**, **[shuttle-3-diffusion](https://huggingface.co/shuttleai/shuttle-3-diffusion)**, and choose the LoRA for integration with MoE-LoRA including **[Turbo-Alpha](https://huggingface.co/alimama-creative/FLUX.1-Turbo-Alpha)**, **[AntiBlur](https://huggingface.co/Shakker-Labs/FLUX.1-dev-LoRA-AntiBlur)**, **[Midjourney-Mix2](https://huggingface.co/strangerzonehf/Flux-Midjourney-Mix2-LoRA)**, **[Super-Realism](https://huggingface.co/strangerzonehf/Flux-Super-Realism-LoRA)**, **[Chatgpt-Ghibli](https://huggingface.co/openfree/flux-chatgpt-ghibli-lora)**. Choose the model you like and download it. For the MoE-LoRA, we will open source a unified checkpoint that can be used for both 512 and 1024 resolutions. 

Before executing the script, download **[Qwen3-8B](https://huggingface.co/Qwen/Qwen3-8B)** to select the task type for the input instruction, base model(**FLUX.1-Krea**, **FLUX.1-dev**, **FLUX.1-schnell**, **shuttle-3-diffusion**), **[MLLM](https://huggingface.co/Qwen/Qwen2.5-VL-7B-Instruct)** and **[Alignet](https://huggingface.co/OPPOer/X2I/blob/main/qwen2.5-vl-7b_proj.pt)**. All scripts follow analogous command patterns. Simply replace the script filename while maintaining consistent parameter configurations.

```shell
$ python infer.py --device cuda --pixel 1024 --num_experts 12 --base_path BASE_PATH --qwen_path QWEN_PATH --lora_path LORA_PATH --extra_lora_path EXTRA_LORA_PATH
```

**device:** The device used for inference. default: `cuda`<br>
**pixel:** The resolution of the input image, , you can choose from **[512, 1024]**. default: `1024`<br>
**num_experts:** The number of expert in MoE. default: `12`<br>
**base_path:** The path of base model.<br>
**qwen_path:** The path of model used to select the task type for the input instruction. We use **Qwen3-8B** here.<br>
**lora_path:** The path of MoE-LoRA in X2Edit.<br>
**extra_lora_path:** The path of extra LoRA for plug-and-play. default: `None`.<br>

## Citation

🌟 If you find our work helpful, please consider citing our paper and leaving valuable stars

```
@misc{ma2025x2editrevisitingarbitraryinstructionimage,
      title={X2Edit: Revisiting Arbitrary-Instruction Image Editing through Self-Constructed Data and Task-Aware Representation Learning}, 
      author={Jian Ma and Xujie Zhu and Zihao Pan and Qirong Peng and Xu Guo and Chen Chen and Haonan Lu},
      year={2025},
      eprint={2508.07607},
      archivePrefix={arXiv},
      primaryClass={cs.CV},
      url={https://arxiv.org/abs/2508.07607}, 
}
```